Jack Dorsey Fired 4,000 Block Workers for AI. Then the Rehires Started.

On February 26, 2026, Jack Dorsey told Block's investors that artificial intelligence had made roughly 4,000 of his 10,000 employees unnecessary. The stock jumped 24 percent the same day. Six weeks later, Richard Hesse, a technical lead at the company, told his manager he would quit unless they rehired members of his team. They did.

That's the story Block doesn't want to lead with. Hesse said he was the only person left running infrastructure that customers depended on. According to Yahoo Finance, at least four employees have been brought back so far. The company described some of the original cuts as the result of "clerical and administrative errors." Bloomberg published a separate report in early March suggesting parts of the layoff were closer to AI-washing than AI substitution.

This week the story got an extra layer. On April 17, Fortune published a long interview with Dorsey explaining how the decision was made. He described running an internal exercise at the end of 2025 to calculate "the minimal number of people that we would need to keep the service up 100 percent." He admitted the calculation included mistakes. He also said he believes most other companies will reach the same conclusion within a year, whether they want to or not.

The market is paying attention. Block trades around $68 per share, with a market cap near $52 billion as of mid-April, up from roughly $41 billion before the layoffs. Analysts have raised price targets. The company raised its own 2026 gross profit guidance to $12.2 billion. From a pure stock price perspective, firing 4,000 workers worked. From an operating perspective, the test is just starting.

The Numbers

Block reported Q4 2025 gross profit of $2.87 billion, up 24 percent year over year. Full-year 2025 gross profit was $10.36 billion, up 17 percent. The company expects $12.2 billion in gross profit for 2026, an 18 percent annual increase. Management guided to $3.2 billion in adjusted operating income at a 26 percent operating margin.

Cash App did most of the heavy lifting. Its gross profit grew 24 percent. Revenue grew 33 percent, driven by what the company called higher user engagement and conversion. Cash App commerce enablement volume hit $54.7 billion in Q4 2025, up 17 percent. Square, the older business, grew gross payment volume 10.3 percent in Q4 and accelerated past 12 percent through February 24, 2026.

Then came the headcount cut. Block went from over 10,000 employees to under 6,000. Severance for affected workers was 20 weeks or more depending on tenure, plus equity vesting through end of May, six months of health coverage, the corporate device, and an extra $5,000. By the standards of tech layoffs, the package was generous. By the standards of stable employment at a profitable $52 billion company, it was sudden.

Stock analysts liked it. According to MarketBeat, the consensus 12-month price target sits around $80, with some forecasts as high as $131. Public.com lists 32 analysts at a Buy consensus as of April 15. The bull thesis is straightforward: Cash App keeps growing, Square reaccelerates, and the smaller workforce drops billions of dollars in compensation expense to the operating line. The bear thesis is that the company has not yet operated through a full quarter at the new headcount, and the rehires suggest the original math missed something.

Pressure Points

The Operational Floor

Dorsey's framing was that Block measured the minimum staffing needed to keep services running at 100 percent. The phrase carries a hidden assumption. It assumes Block leadership knew which employees were doing essential work and which were not. The Hesse situation suggests they did not.

Hesse was a technical lead at Block. He went public with the fact that he had to threaten resignation to get colleagues rehired who maintained customer-facing infrastructure. If a senior technical lead had to escalate to keep his team intact, the original reduction was not a calculation. It was a guess.

This matters because Cash App and Square are not products in the traditional sense. They are real-time financial systems with regulatory exposure. Outages cost money directly, through transaction failures, and indirectly, through compliance scrutiny. Cash App processes payments for tens of millions of monthly users. Square handles point-of-sale transactions for small businesses that have no Plan B if the terminal goes down. The minimum operating headcount is not a fixed number. It is whatever number prevents the next outage.

The rehires are a small data point right now. The signal to watch is whether the rehire count grows quietly through the summer, and whether Block discloses it.

The Productivity Story Has to Hold

Dorsey's pitch was not just cost reduction. It was that AI tools had reached a level where Block could be smaller and produce the same output, or more. He named specific tools in the Fortune interview. Anthropic's Opus 4.6. OpenAI's Codex 5.3. The argument is that engineering and operational tasks that previously required teams now require individuals with AI assistance.

That argument has to show up in the numbers within four quarters. Specifically, it has to show up as gross profit per employee climbing materially without a corresponding deterioration in product velocity, customer satisfaction, or compliance posture. If Block ends 2026 with the promised $12.2 billion in gross profit and roughly 6,000 employees, that is about $2 million in gross profit per employee, up from about $1 million in 2025. That doubling is the proof point.

If the number lands, Dorsey's bet looks correct, and other CEOs follow. If gross profit per employee comes in at $1.4 to $1.6 million instead, and product complaints rise, the story becomes a cautionary case study about how AI tools were not yet ready to absorb that much human work.

Bloomberg's reporting on AI-washing matters here too. The pandemic-era hiring at Block was substantial. The company added thousands of roles between 2020 and 2022, many in product and engineering, some of which probably should have been trimmed regardless of AI. Bundling that overdue cleanup with an AI narrative makes the cost cut sound visionary instead of overdue. Investors do not seem to mind. Regulators and labor analysts will.

The Hiring Pipeline

Block's compensation has historically been competitive but not at the very top of fintech. The company recruited talent partly on mission, partly on Dorsey's personal brand, partly on the chance to work on payments at scale. The 40 percent layoff changes the recruiting equation.

The next senior engineer Block tries to hire is going to ask why their predecessor was let go. The answer matters. If the company says some roles were AI-substituted, candidates will calculate the probability that their role gets the same treatment in 18 months. If the company says it was a temporary correction, candidates will discount the equity offer accordingly. Either way, the cost of attracting top talent goes up.

The boomerang trend matters here too. According to AZFamily and other outlets covering the broader pattern in early April 2026, multiple companies that did AI-driven layoffs in late 2025 and early 2026 are now rehiring some workers, often at lower seniority and tighter compensation. That trend is being tracked by labor researchers, and it shapes how candidates negotiate. The market knows this game can reverse.

What Happens Next

The most likely path: Block delivers gross profit roughly in line with guidance for the first half of 2026, somewhere between $5.7 and $5.9 billion. Cash App growth holds. Square keeps reaccelerating. Operating margin expands to the high 20s. The rehire count grows quietly to maybe 50 or 100 employees, almost all in compliance, infrastructure, and customer support. Dorsey's bet looks vindicated through the summer.

Then a stress test arrives. A Cash App outage that hits during peak retail traffic. A regulator asking why anti-money-laundering staffing dropped so quickly. A Square integration that breaks during a major payments season and takes longer than expected to fix because the team that owned it no longer exists. Whether any of these happens is uncertain. The probability that none of them happens within 12 months is low.

The bull case: AI tools genuinely close the gap. Block ships product faster than ever in the third and fourth quarters. Stock pushes through $100. Other CEOs follow Dorsey's playbook. Block becomes the case study in what an AI-native company looks like, and Dorsey gets credit for making the call early. This requires the productivity claim to be real, not just a story.

The bear case: A serious outage or regulatory action in the second half of 2026 forces Block to disclose substantial rehiring. The stock corrects 30 to 40 percent on doubts about the operating model. Other companies that copied Dorsey end up in the same position. The phrase "AI-washing" enters the SEC filings vocabulary. By late 2026, the conversation shifts from "AI replaces workers" to "AI reshapes work." Different stories. Different valuations.

What To Watch

Five signals over the next two quarters:

• Gross profit per employee in Q1 and Q2 2026 reports. The doubling claim has to show up. Anything below $1.6 million per employee on an annualized basis is a yellow flag. Anything below $1.4 million is a red flag.
• Rehire disclosures. Block has not committed to publishing rehire numbers. If a journalist or labor researcher gets the count past 200, that becomes a credibility problem for the original communication.
• Cash App active user growth. Customer support quality has a habit of showing up in retention metrics. If monthly transacting actives stall, the AI-substitution thesis weakens fast.
• Senior engineering departures. Watch for VPs and directors leaving for competitors. A wave of senior departures within six months would suggest the cuts hit institutional knowledge harder than the operating model assumed.
• Other CEO announcements following Dorsey's playbook. If Salesforce, Workday, or a major bank announces a similar 30 to 40 percent cut citing AI, the signal becomes a market-wide trend. If they don't, Block is the outlier, and the experiment becomes easier to evaluate in isolation.

My Opinion

Dorsey is doing two things at once, and they should be evaluated separately. The first is making a real bet that AI tools have changed the production function for software and operations. That bet might be right. There is genuine evidence that current models can shoulder meaningful coding, support, and analytics work. If you assume the technology keeps improving, smaller teams producing more output is a real possibility.

The second thing he is doing is using that bet to make a workforce decision faster than the evidence supports. The Hesse story is small but instructive. A senior technical lead had to threaten to quit to keep customer infrastructure staffed. That is not a measured reduction. That is a cut first, sort it out later. The company is now sorting it out, employee by employee, in a way that does not show up in press releases.

The deeper issue is honesty about what AI can and cannot do. Dorsey is telling other CEOs to follow him. He is also quietly admitting that the 40 percent number was wrong in some cases. Both can be true. But the public confidence and the private corrections need to converge into a clearer story before the next round of CEOs makes the same call. Right now they are not converging. They are running in parallel, and the gap is what investors, employees, and regulators will eventually have to price.

If Block makes its 2026 numbers, Dorsey wins this argument and the AI labor reorganization accelerates across the S&P 500. If Block misses, or worse, if a customer-facing failure exposes the rehire trail, the conversation snaps back to where it was a year ago: AI as augmentation, not replacement. The most interesting fact about this story is that we will know which one is true within four quarters. Few corporate experiments come with a verdict that fast.